Fossil Evidence Suggests Giant Octopuses Dominated Oceans 100 Million Years Ago

A team led by Yasuhiro Iba of Hokkaido University has published fossil evidence in Science suggesting that octopuses reaching up to 19 metres in length were apex predators during the Late Cretaceous — a claim that, if it holds, rewrites our understanding of who ruled the ancient oceans [1][2]. The study arrives at a moment when cephalopod paleontology is already in flux: just two weeks earlier, a separate paper reclassified the "world's oldest octopus" fossil as a nautilus relative [3]. The field is both finding and losing ancient octopuses at the same time.

But how solid is the evidence? And does "fossil evidence suggests giant octopuses" really translate to "dominated oceans"?

The Fossils: 27 Jaws and a New Technique

The core evidence consists of 27 fossilized cephalopod jaws — the chitinous beak structures that are among the only hard parts an octopus possesses [1][2]. Fifteen of these were previously known specimens from collections in Japan and Vancouver Island, Canada. The remaining 12 were newly discovered using a technique Iba calls "digital fossil mining": high-resolution grinding tomography, in which rock samples are ground down layer by layer and photographed at each step, with a zero-shot learning AI model identifying jaw fossils embedded inside [4][5].

The fossils span 100 to 72 million years ago, placing them firmly in the Late Cretaceous [1]. They were recovered from fine-grained seafloor sediments — the kind of calm depositional environment that preserves delicate organic structures [4]. The specimens are attributed to two species within the genus Nanaimoteuthis: the larger N. haggarti and the smaller N. jeletzkyi, estimated at roughly 8 metres [2][6].

Critically, these are finned octopuses (vampyromorphs or close relatives), not the benthic octopuses most people picture. Modern finned octopuses, like the dumbo octopus, are deep-water animals with a very different body plan from shallow-water species [6].

The Size Estimate: How 19 Metres Was Derived

The headline figure — 19 metres, or about 62 feet — comes from scaling relationships between beak size and total body length in living finned octopuses [2][6]. The largest N. haggarti beak was approximately 1.5 times larger than the beak of a modern giant squid [6]. Applying the beak-to-body ratio observed in extant relatives yields a range of 7 to 19 metres in total length [2].

That range matters. The lower bound (7 metres) describes an animal about the size of a large modern giant Pacific octopus — impressive, but not unprecedented. The upper bound (19 metres) describes something longer than most mosasaurs. The published estimate effectively spans a factor of nearly three.

Iba has stated that the findings hold "even with a wide margin of error" [5], and Christian Klug, a cephalopod paleontologist at the University of Zürich who was not involved in the study, said he is "pretty sure" about the larger estimate — but also called the 19-metre figure "quite extreme" [6]. Klug noted that the relationship between mantle length and tentacle length in modern cephalopods is highly variable, meaning the lower or middle bounds of the range may be more probable [6].

No body mass estimate has been published. Without soft tissue, estimating mass from beak size alone is far more uncertain than estimating length, and the authors appear to have avoided that calculation.

The Predator Question: Could an Octopus Compete with Mosasaurs?

The study reports significant wear on the largest jaws — up to 10% of the jaw tip had been ground away relative to total jaw length, exceeding wear patterns in any modern hard-prey-feeding cephalopod [4]. The wear includes scratches, chips, and rounded edges consistent with repeatedly crushing shells and bones [1][2]. This is the primary evidence for an apex predatory role.

Source: Compiled from published estimates (Iba et al. 2026, Everhart 2005, O'Keefe & Carrano 2005)

Data as of Apr 23, 2026CSV

The Late Cretaceous ocean was already crowded with large predators. Mosasaurs like Mosasaurus hoffmannii reached 17 metres; Tylosaurus proriger reached about 14 metres; large plesiosaurs like Elasmosaurus stretched to similar lengths [7][8]. These were vertebrates with bones, teeth, and — in the case of mosasaurs — muscular tails adapted for fast pursuit. An octopus, even a giant one, would have had a fundamentally different predatory strategy: ambush rather than chase, intelligence and camouflage rather than speed and bite force.

Whether a 19-metre octopus could have preyed on marine reptiles remains speculative. The jaw evidence shows it ate hard-shelled organisms. Whether those included the bones of vertebrates, as some coverage has implied, is not established beyond the presence of bone-like wear patterns [4]. Adiel Klompmaker of the University of Alabama noted that deeper ocean specimens — where these animals likely lived — remain almost entirely unsampled [6].

The study's actual claim is more restrained than much of the media framing: that these octopuses were "among the top ocean predators," not that they dominated the entire ecosystem [1][4]. The paper provides "the first direct evidence that invertebrates could evolve into giant, intelligent apex predators in ecosystems that have been dominated by vertebrates for about 400 million years" [4].

The Fossilization Problem: Why Skepticism Is Built In

Octopus fossils are extraordinarily rare. The animals lack bones, internal shells (in most lineages), and mineralized structures. The only durable parts are the chitinous beaks and, occasionally, ink sacs [9][10]. The entire known octopus fossil record amounts to a few dozen specimens spanning hundreds of millions of years.

Source: Compiled from Klug et al. 2015, Kruta et al. 2016, Iba et al. 2026

Data as of Apr 23, 2026CSV

This rarity means every major octopus fossil claim receives intense scrutiny — and several have not survived it. The most prominent recent example: Pohlsepia mazonensis, a 300-million-year-old fossil that held the Guinness record as the oldest known octopus, was reclassified on April 8, 2026 as a decomposed nautiloid relative. Researchers at the University of Reading found it possessed a radula with at least 11 teeth per row, where octopuses have only seven or nine [3][11]. The decomposition that preserved it was precisely what had made it look octopus-like.

Mark McMenamin's 2011 "Triassic kraken" hypothesis — that a giant octopus arranged ichthyosaur bones into a self-portrait — was never published in a peer-reviewed journal and was widely dismissed as closer to cryptozoology than paleontology [12][13].

The Iba et al. study clears a meaningfully higher bar than either of these cases. It was published in Science, one of the two most selective general-science journals, after peer review [1]. It relies on physical specimens (27 jaws) rather than indirect proxies like bone arrangements. And the taxonomic identification is based on detailed morphological comparison with known cephalopod jaw types, not on ambiguous soft-tissue impressions [4][6].

That said, Misha Whittingham, a postdoctoral researcher at the Open University in the UK, emphasized that octopus evolution remains "largely a mystery, since so few of their fossils are ever found" [5]. The study pushes back the oldest known octopus fossils by approximately 5 million years — but each such revision also underscores how incomplete the record is [2].

What Drove Cretaceous Gigantism?

The mid-Cretaceous ocean was a different world. Sea surface temperatures in tropical waters reached 32–36°C, compared to roughly 27°C today [14]. Atmospheric CO₂ levels were likely 1,000–2,000 ppm, versus about 425 ppm in 2026 [15]. These conditions drove oceanic anoxic events — episodes of widespread oxygen depletion that reshaped marine ecosystems [14][15].

Higher temperatures and lower oxygen would seem to constrain, not enable, gigantism. Warmer water holds less dissolved oxygen, and larger animals require more of it. The metabolic modeling for a 19-metre octopus in these conditions has not been published, and the Iba et al. paper does not address it directly [4].

One possibility: finned octopuses, which today inhabit deep, cold, oxygen-rich waters, may have exploited similar deep-water refugia during the Cretaceous. The fossils come from fine-grained sediments consistent with deep or quiet-water environments [4][5]. If the deep ocean remained cooler and better oxygenated than the surface — as some paleoceanographic models suggest [16] — then gigantism in a deep-dwelling cephalopod becomes more plausible, though still unmodeled.

The prey base question is similarly open. A 19-metre predator requires a substantial food supply. The Late Cretaceous ocean supported diverse ammonite, fish, and shark populations, but quantitative estimates of available biomass at depth are not available for this interval.

The Evolutionary Bottleneck: From 19 Metres to 4

If giant octopuses truly existed during the Cretaceous, something eliminated them. The Cretaceous-Paleogene (K-Pg) extinction event 66 million years ago is the obvious candidate. The asteroid impact that killed non-avian dinosaurs also devastated marine ecosystems: ammonites went entirely extinct, and marine reptiles were wiped out [17].

Recent genomic studies confirm that modern decapodiform cephalopods (squid, cuttlefish) underwent a severe evolutionary bottleneck at the K-Pg boundary, with survivors likely sheltering in deep-ocean refugia with less severe acidification and temperature disruption [17][18]. A similar or more extreme bottleneck for octopuses is consistent with the available evidence, though direct fossil confirmation is lacking.

Source: OpenAlex

Data as of Jan 1, 2026CSV

Today, the largest octopus species — the giant Pacific octopus — reaches about 5 metres arm-span [2]. The largest living cephalopod, the colossal squid, may reach 10–12 metres. Neither approaches the upper estimate for N. haggarti. Academic interest in Cretaceous cephalopod fossils has grown steadily, with 179 papers published on the topic in 2023 alone, reflecting increasing recognition that the cephalopod fossil record contains major gaps [19].

The size collapse from the Cretaceous to the modern era could reflect the loss of the specific ecological conditions — deep-water prey abundance, absence of competition from marine mammals — that permitted extreme size. After the K-Pg event, marine mammals eventually radiated into many of the niches formerly held by marine reptiles and, perhaps, giant cephalopods.

The Steelman Case for Overstatement

The gap between the paper's claims and the headlines deserves direct examination. The study demonstrates that large octopuses existed during the Late Cretaceous and occupied predatory roles. It does not demonstrate "dominance."

"Dominated oceans" implies population-level ecological control — that these animals were abundant, widespread, and shaped the structure of marine food webs. The evidence consists of 27 jaws from two locations (Japan and Vancouver Island) spanning a 28-million-year window [1][4]. That is consistent with a rare, large predator — analogous to how sperm whales are apex predators today without "dominating" the ocean.

The "kraken" framing, used by multiple outlets and arguably encouraged by the study's own language about "top predators," evokes a mythological sea monster [6][2]. This is effective science communication in that it generates public interest, but it also risks conflating a carefully bounded fossil finding with a population ecology claim that the data cannot support.

Christian Klug's measured assessment — that the upper size estimate is "quite extreme" but probably in the right range — is arguably the most honest summary [6]. These were probably large, possibly very large, predatory octopuses. Whether they were 7 metres or 19 metres, and whether they were rare deep-water specialists or abundant apex predators, remains genuinely uncertain.

Who Is Behind the Research?

The study was led by Yasuhiro Iba, a professor in the Department of Earth and Planetary Sciences at Hokkaido University in Japan, with collaborators from Japan and Germany [4][5]. The lead author on the paper is listed as Shin Ikegami et al. [4]. The fossils from Canada are housed at the Courtenay and District Museum and Paleontological Centre in British Columbia [5].

The paper was published in Science (DOI: 10.1126/science.aea6285) on April 23, 2026, meaning it underwent the journal's full peer review process [1]. Independent experts quoted in coverage include Christian Klug (University of Zürich), Adiel Klompmaker (University of Alabama), Cameron Tsujita (Western University, Ontario), and Misha Whittingham (Open University, UK) [5][6]. None of these external scientists have reported examining the physical specimens directly; their assessments are based on the published data.

What This Changes — and What It Doesn't

The Iba et al. study is a genuine contribution. It adds 12 new specimens to an almost nonexistent fossil record, demonstrates a novel AI-assisted discovery technique, and provides the strongest evidence yet that Cretaceous octopuses were substantially larger than their modern descendants. The peer review in Science lends it credibility that previous giant cephalopod claims — the Triassic kraken, the misidentified Pohlsepia — never achieved.

But the distance between "some Cretaceous octopuses were large predators" and "giant octopuses dominated the oceans" is vast, and it is a distance the fossil evidence does not currently cross. Twenty-seven jaws from two sites do not establish population-level dominance. The size range of 7–19 metres is wide enough that the animal could have been impressive or extraordinary. And the metabolic and ecological modeling needed to support a 19-metre soft-bodied predator in warm, low-oxygen Cretaceous waters has not yet been done.

The finding is significant on its own terms. It does not need the "kraken" mythology to be interesting.